Just playing with it, using the tutorial program, regardless of what the reviews say, I'd argue, that, as long as you are proficient in the use of it, and tapping on a consistent surface (i.e. you're sitting or standing still, with your knee reasonably motionless), it should be fine. It makes it clear in the tutorials, something many reviewers seem to either disregard, or are incapable of executing, that light taps are highly preferable. Makes sense if you consider the mechanism, heavier vibrations are more likely to trigger the other sensors, and it relies on being able to differentiate singular or combination triggers.
It won't be an easy learning curve, I personally believe this product would be more adapted to two hand use with a second input vector added as a shift/function key at a minimum, and think it could easily add a couple more input axis, however, for your intended use, if you're willing to put in the work and you get a good unit, I could easily see the advantages. Honestly, considering the illegible scrawl I've seen of my doctors, what are a few typo's in the broader picture?
All that said, while I'm highly hopeful, and see a lot of untapped potential for this product personally, I'd recommend being very cautious in the assumption that this product will continue to exist, let alone, refine, in the future. I certainly hope so, though it seems dependent on the developer, or it's tech, being acquired, and fostered, by a larger player, or a larger iterative funding round, with direction and refinement, moving forward. It's highly promising, with more potential than I think many people understand, but needs improvement, which further vision to penetrate the highly stagnant and entrenched ethos of so-called computer input devices.